AC133 Antigen Expression Is Not Restricted to Acute Myeloid Leukemia Blasts But Is Also Found on Acute Lymphoid Leukemia Blasts and on a Subset of CD34⁺ B-Cell Precursors

To the Editor:

Human AC133 antigen is a 5-transmembrane molecule that belongs to the prominin family.1-4 In the hematopoietic system, AC133 expression is restricted to a subset of CD34⁺progenitor/stem cells1 and to leukemic blasts from patients with acute myeloid leukemia (AML).2,5-7 Controversy exists whether the AC133 antigen is also expressed on acute lymphoblastic leukemia (ALL) blasts. One group described low levels of AC133 antigen expression on the blasts of some ALL samples,2 whereas 3 other groups reported the exclusive expression of this molecule on AML blasts.5-7 Using multiparameter flow cytometry, we analyzed the correlated expression of AC133 antigen and CD34 in 23 AML and 9 ALL samples from adult patients, as well as on the surface of normal CD34⁺ bone marrow cells. Table1 shows the distribution of AC133 antigen and CD34 on leukemic blasts in correlation to their morphological classification. In line with previous reports, the majority of CD34⁺ AML samples (15/18 [83%]) coexpressed the AC133 antigen. Four of the 5 tested CD34⁻ AML samples (80%) were also AC133⁺. Notably, these samples exclusively displayed the M₄/M₅ French-American-British (FAB) subtypes. Four of 10 CD34⁺AC133⁺ AML samples of the M₁/M₂ subtypes and 1 of 6 of the M₄/M₅ subtypes expressed high levels of AC133 antigen (mean log fluorescence intensity between the second and third decade). In addition, 1 of 4 CD34⁻AC133⁺ AML samples were also AC133^bright. Representative plots of a CD34⁺AC133^bright FAB M₂ and a CD34⁻AC133^dim FAB M₄ sample are shown in Fig 1. In contrast to other reports, our data neither confirm that AC133^bright cells are exclusively found in M₄/M₅ AML subtypes2 nor that AC133^bright blasts resemble exclusively the most immature FAB types,7 but rather suggest that high levels of AC133 antigen expression is observed on myeloid leukemic blasts of all FAB subtypes.

Table 1 also demonstrates that 3 of 6 C-ALL, 1 of 1 pro-B ALL, and 1 of 2 T-ALL samples were AC133⁺. These data are in line with the observations of Miraglia et al,2 who found AC133 antigen expression on 4 of 6 ALL samples, but do not support the findings of Snell et al,7 who could not detect AC133 antigen expression on any of the 17 tested ALL samples. In contrast to Miraglia et al,2 we found high levels of AC133 expression in some B-lymphoid ALL samples (2/7). Figure 1 shows that 1 AC133^bright sample was of the C-ALL and 1 was of the pro-B ALL phenotype. Notably, the AC133^bright blasts of the pro-B ALL sample were heterogeneous with regard to CD34 expression. However, whether the CD34⁺AC133^bright and the CD34⁻AC133^bright represent populations of different maturational stages has yet to be determined.

To analyze whether AC133⁺ B-lymphoid cells exist also among nonleukemic hematopoietic precursor cells, normal bone marrow cells enriched for CD34⁺ cells by MACS technology were stained with anti-CD34-peridine chlorophyll protein (PerCP), AC133-phycoerythrin (PE), and anti-CD10-fluorescein isothiocyanate (FITC) or anti-CD19-FITC, respectively, and analyzed on a FACSCalibur flow cytometer. Figure 2 shows that the majority of the CD34⁺CD10⁺ and CD34⁺CD19⁺ cells (34.8% v 29.4%) are AC133⁻. However, a minor CD34⁺CD10⁺ and CD34⁺CD19⁺ B-cell precursor population exists (15.1% v 5.9%) that is also AC133⁺ and may represent the normal counterpart of AC133⁺ ALL cells. Our data suggest that AC133 antigen is not only expressed on stem cells and myeloid progenitor cells and their leukemic counterparts,1 2 but also on ALL blasts and normal CD34⁺ B-lymphoid precursor cells.